Calibrating polypropylene particle model parameters with upscaling and repose surface method

Abstract

The discrete element method (DEM) is a computational technique extensively utilized for simulating particles on a large scale, specifically focusing on granular materials. Nonetheless, its implementation requires a substantial amount of computational power and accurate material properties. Consequently, this study delves into an alternative approach referred to as volume-based scaled-up modeling, aiming to simulate polypropylene particles using DEM while mitigating the computational burden and regenerating new material properties. This novel method aims to reduce the CPU time required for the simulation process and represent both the macro mechanical behavior and micro material properties of polypropylene particles. To accomplish this, the dimensions of the polypropylene particles in the DEM simulation were magnified by a factor of two compared to the original size of the prolate spheroid particles. In order to determine the virtual micro material properties of the polypropylene particles, a calibration method incorporating the design of experiments (DOE) and repose surface methodology was employed. The predicted bulk angle of repose (AOR) derived from the upscaled DEM parameters exhibited a remarkably close agreement with the empirical AOR test, demonstrating a small relative error of merely 1.69 %. Moreover, the CPU time required for the upscaled particle model proved to be less than 71 % of that necessary for the actual-scale model of polypropylene particles. These compelling results confirm the effectiveness of enlarging the particle volume used to calibrate micro-material properties in the Discrete Element Method (DEM) through the DOE technique. This approach proves to be a reliable and efficient method